Table 3 Class-wise performance of ngLOC method on eukaryotic datasets

Cytoplasm

CYT

0.818

0.750

0.983

0.762

0.864

0.832

0.991

0.838

Cytoskeleton

CSK

0.937

0.784

0.998

0.853

0.988

0.965

1.000

0.976

Endoplasmic Reticulum

END

0.970

0.785

0.999

0.869

0.876

0.645

0.999

0.748

Extracellular

EXC

0.953

0.946

0.974

0.922

0.966

0.723

0.999

0.831

Golgi Apparatus

GOL

0.940

0.593

1.000

0.745

1.000

0.509

1.000

0.712

Lysosome

LYS

0.949

0.693

1.000

0.810

Mitochondria

MIT

0.979

0.852

0.998

0.906

0.912

0.727

0.995

0.804

Nuclear

NUC

0.805

0.914

0.960

0.831

0.769

0.873

0.976

0.802

Plasma Membrane

PLA

0.876

0.957

0.961

0.890

0.796

0.866

0.989

0.822

Peroxisome

POX

0.946

0.760

1.000

0.847

0.906

0.580

1.000

0.724

Cell Junction

JNC

0.774

0.387

1.000

0.547

Chloroplast

CHL

0.946

0.977

0.899

0.889

Vacuole

VAC

0.844

0.702

0.998

0.766

% Overall accuracy

89.88%

91.39%